sign in sign up
<<
Home , Brown palm civet, Cullenia, Cullenia exarillata, Elaeocarpus, Elaeocarpus serratus, Ficus amplissima

Journal of Mammalogy, 89(6):1550–1556, 2008

FOOD HABITS OF THE INDIAN GIANT FLYING SQUIRREL (PETAURISTA PHILIPPENSIS) IN A RAIN FOREST FRAGMENT,

R. NANDINI* AND N. PARTHASARATHY Department of Ecology and Environmental Sciences, Pondicherry University, Puducherry, 605 014,

Present address of RN: National Institute of Advanced Studies, Indian Institute of Science campus, Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021 Bangalore, 560 012, India Present address of RN: Department of Biological Sciences, Auburn University, Auburn, AL 36849, USA

We examined the feeding habits of the Indian giant flying squirrel (Petaurista philippensis) in a rain-forest fragment in southern Western Ghats, India, from December 1999 to March 2000. Flying squirrels consumed 4 major parts belonging to 9 plant . racemosa was the most-eaten species (68.1%) during the period of the study, followed by exarillata (9.57%) and Artocarpus heterophyllus (6.38%). The most commonly consumed food item was the of F. racemosa (48.93%). formed an important component of the diet (32.97%) and the leaves of F. racemosa were consumed more than those of any other species. Flying squirrels proved to be tolerant of disturbance and exploited food resources at the fragment edge, including exotic planted species.

Key words: edge, Ficus, fig , folivore, Petaurista philippensis, rain-forest fragment, Western Ghats

The adaptability of allows them to exist in varied across the Western Ghats seem to increase with disturbance. environments and helps them to cope with habitat fragmenta- Ashraf et al. (1993) found that P. philippensis was encountered tion. Predicting the reaction of mammals to landscape most in cardamom plantations, and Umapathy and Kumar modification and fragmentation is challenging because various (2000) reported that densities increased with disturbance in species react differently to the same perturbations (Johns 1997). rain-forest fragments in the Anamalai Hills of the Western Even closely related species that share life-history strategies Ghats. Similar trends have been seen for other species of the and dietary characteristics react divergently (Simberloff and Petaurista in Southeast Asia, where population densities Dayan 1991; Thiollay 1992). Some species are known to of Petaurista were higher in logged forests than in unlogged increase in density after habitat fragmentation, whereas others forests (Barrett 1984). decline or disappear. Species that are persistent are often The Western Ghats, a biodiversity hotspot (Myers et al. otherwise known to be adaptable, and are able to modify their 2000) in southern India, is under increasing human pressure behavior, ranging patterns, and dietary patterns to accommo- with a forest-cover loss of 25.6% over the past 25 years (Jha date changes in forest structure and composition (Johns 1997). et al. 2000). Conversion of forested areas to plantations that Feeding ecology is an important aspect of the survival ability destroy or alter the habitat is known to affect the population of a species. The forest structure, plant productivity patterns, size and dispersal ability of flying squirrels in temperate forests and availability of resources change after disturbance, and and result in the invasion of competitors, predators, and some herbivores cope by exploiting common resources, pathogens (Weigl et al. 1999). whereas others reduce activity and conserve energy (Meijaard Muul and Lim Boo Liat (1978) described large flying et al. 2005). squirrels of the genus Petaurista as being primarily folivorous The Indian giant flying squirrel (Petaurista philippensis)is based on their stomach morphology. As part of a larger study to a nocturnal arboreal whose populations in some sites examine effects of habitat fragmentation on flying squirrels, we conducted a preliminary examination of the diet of flying squirrels in a rain-forest fragment in the Anamalai Hills of the * Correspondent: [email protected] Western Ghats for a period of 3 months. Our objectives were to examine if flying squirrels use food items in proportion to Ó 2008 American Society of Mammalogists availability or if they are selective in their use of resources; to www.mammalogy.org examine the use of introduced plant species in plantations and 1550 December 2008 NANDINI AND PARTHASARATHY—DIET OF PETAURISTA IN FRAGMENTED RAIN FOREST 1551 forest fragments in the diet of flying squirrels; and to examine the nearest (30 cm girth at breast height) in each of the use of the forest fragment interior and adjoining habitats by 4 quadrats, and the girth, height, and species identification flying squirrels. of these were recorded. The presence of understory in increasing height-class intervals was recorded around the focal tree. MATERIALS AND METHODS Feeding habits of P. philippensis.—Trails were walked in Study site.—The study was conducted in a rain-forest the forest interior and edge between 1830 and 2230 h. Flying fragment near Indira Gandhi Wildlife Sanctuary, southern squirrels were detected by eyeshine and the sound of their Western Ghats. This sanctuary, one of the largest protected calls or occasionally by their movement in the trees. Flashlights areas in southern India, encompasses rain-forest fragments (12 V) were used and observations were made using a pair of ranging from 1 to 2,500 ha in size (Prabhakar 1998). Most Nikon 7 50 binoculars (Nikon, Inc., Melville, New York). Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021 fragments in and around the sanctuary were created between Once a squirrel was identified, the tree species it was on, tree 1860 and 1978 when the forests were leased for the cultivation height, height of the on the tree, and behavior of the of tea, coffee, cardamom, Eucalyptus, rubber, and teak. The animal were recorded. When a squirrel was seen feeding, the shortest distance between any fragment and the nearest forest plant part consumed and the phenophase of the part were noted. patch ranges between 1 and 5 km. This method of diet calculation emphasizes the diversity of The study site, Puthuthottam, is located on the Valparai various items ingested by the squirrel, but provides no data on plateau (;220 km2), which comprises rain-forest fragments the amount of each species consumed (Paschoal and Galetti amid privately owned tea, coffee, Eucalyptus, and cardamom 1995). Once the observation was made, we moved along the plantations. Puthuthottam (108209N, 768589E, elevation 1,085 m), trail until the next individual was sighted. Squirrels did not 92 ha in size, was formed in 1906 (Congreve 1938), and is seem to flee the light, especially when they were engaged in the isolated, surrounded by tea and coffee plantations with no act of feeding. Only the 1st few seconds of observation of an corridor linking it to other fragments. The nearest large frag- animal’s behavior were used to collect data in the manner of an ment, the Iyerpadi–Akkamalai complex (;2,000 ha) is 4 km instantaneous scan sample. However, a few individuals did away. The fragment was under-planted with coffee and carda- seem to be bothered by the light, and if they altered their mom and has a history of selective logging, the last time being behavior before the observer could ascertain their initial in 1992 (Prabhakar 1998). Being close to human habitation, it behavior, the data point was discarded during analysis. Few is highly disturbed with regular felling of trees and collection data were rejected through such disturbance caused by the of fuel wood. observer. The area receives ;3,000 mm of rain annually, and the Vegetation parameters were quantified for the squirrel’s food vegetation is classified as western tropical wet evergreen forest trees as well as random trees to provide a comparison of food- (West Coast tropical evergreen forest; climax vegetation tree choice against the available habitat. The parameters type—Mesua–Cullenia– association—Champion recorded were girth at breast height, height, canopy cover and Seth 1968). Despite its small size and isolation, the and canopy contiguity of the focal tree, presence–absence of fragment is known to support threatened and restricted-range surrounding understory, and the distance to the 4 nearest trees arboreal mammals (Rajamani et al. 2002; Umapathy and and their girths. Kumar 2000). Fieldwork was carried out between December Encounter rates.—In addition to the above-mentioned trails, 1999 and March 2000, and coincided with the dry season. we also walked trails in the plantations to estimate the Estimation of vegetation parameters.—Vegetation charac- abundance of flying squirrels. The encounter rate was ters were recorded to determine details of the available habitat calculated as the number of flying squirrels seen per kilometer in the fragment. Two habitat types were identified: forest of trail walked. interior and edge (habitat between forest and plantations). The Data analysis.—The density and dominance of tree species site had existing trails around the edge and through the interior, at the site were calculated using the software package and for this study we identified 3 trails that together covered KREBSWIN version 3.1 (Krebs 1998). SPSS (Norusis 1990) most of the fragment. No fresh trails were cut through the forest was used for analysis and to test for significance of the given its already fragmented condition. Also, the site is variables. privately owned, and it was logistically difficult to cut fresh trails. Thirty-four point quarters (Cottam and Curtis 1956) were laid along trails through the interior and along the edge to RESULTS characterize the general habitat. Plots were laid 10 m to the left Forest structure and composition.—The edge and interior and right of the trail alternatively at 100-m intervals along the habitats proved to be different in tree species composition trail. Girth at breast height, height, canopy cover, and canopy (Table 1) and vegetation characteristics (Table 2). The density contiguity of the focal trees were measured. Tree height was of trees varied significantly between the 2 habitats, being initially measured with a clinometer, and visual estimates were higher in the interior than at the edge (Kolmogorov–Smirnov 2- made only once the observer was reasonably certain of sample test, z ¼ 0.72, P , 0.0001). Canopy cover also was accurately estimating height. Canopy-cover readings were significantly greater in the interior (Kolmogorov–Smirnov 2- taken with a densiometer. Around each tree, the distance to sample test, z ¼ 1.41, P ¼ 0.03; Table 2). 1552 JOURNAL OF MAMMALOGY Vol. 89, No. 6

TABLE 1.—Plant species in the diet of Petaurista philippensis and the parts consumed in Puduthottam, a rain-forest fragment in the Western Ghats (December 1999–April 2000).

Part No. No. Total % Species consumed Phenophasea observations/phenophase observations/part observations/species observations/species Flower 9 9 9 9.57 Mesua ferrea Bark — 1 1 1 1.06 Eucalyptus Bark — 5 5 5 5.32 Dimocarpus longan Bark — 1 1 5 5.32 imm 2 4 Leaf uid 2

Artocarpus heterophyllus Leaf m 1 6 6 6.38 Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021 Leaf uid 5 Ficus callosa Leaf m 1 1 1 1.06 Ficus amplissima Leaf m 1 2 2 2.13 Leaf uid 1 Ficus racemosa Leaf imm 4 18 64 68.09 Leaf m 3 Leaf uid 11 Fruit ur 4 46 Fruit sr 10 Fruit r 7 Fruit ur & sr 3 Fruit ur & r 4 Fruit uid 18 Persea macarantha Leaf uid 1 1 1 1.06 Total 94 100

a Codes for phenological phase of plant parts consumed. Leaf: imm ¼ immature, m ¼ mature, uid ¼ unidentified; fruit: ur ¼ unripe, sr ¼ semiripe, r ¼ ripe, ur & sr ¼ unripe and semiripe, ur & r ¼ unripe and ripe, uid ¼ unidentified.

The girth-class distributions of trees in the interior and edge species, M. eminii (18.38%) and Macaranga (10.29%), had were not significantly different (Wilcoxon matched-pairs sign lower basal areas (1.89 m2/ha and 0.32 m2/ha, respectively), rank test, P . 0.05; Fig. 1). However, most trees at the edge indicating that most of the individuals were in the smaller girth (68.75%) were ,100 cm in girth, whereas most trees in the classes (Table 3). interior (55.36%) were 100–200 cm in girth. When both Food habits.—Petaurista philippensis was sighted 322 times habitats are combined, most trees were small (30–60 cm girth), during the course of the study. This is the number of positive and very few trees were large (.350 cm girth). The largest identifications; observations when eyeshine was detected with- trees (450–550 cm; 2.21% of the population) were all restricted out confirmed identification were excluded from the analysis. to the interior. P. philippensis was observed feeding on 103 occasions, and Thirty-four tree species were recorded for the entire site, most observations were at the fragment edge. During the study with 22 in the interior and 16 at the edge. The most abundant period flying squirrels consumed at least 25 different plant parts species in the interior were Maesopsis eminii (23.75%), of 10 tree species from 8 different families (Table 1). There Macaranga (17.50%), Maesa indica (11.25%), and Mesua was 1 observation of a flying squirrel feeding on lichen from ferrea (8.75%; Table 3). M. eminii and Macaranga are pioneer the bark of Macaranga. species, quick to invade natural gaps and cleared areas. Petaurista philippensis consumed fruit more than any other M. eminii is known to self-seed and to develop large soil plant part (44% of the observations; v2 ¼ 81.24, d.f. ¼ 4, seed banks rapidly (Hamilton and Benstead-Smith 1990). The P , 0.001; Fig. 2). Semiripe fruit was consumed (9.80% of high abundance of these species indicates the high level of observations) more than ripe fruit (6.86% of observations; disturbance at the site. Table 1). The only fruits observed being consumed during this The most abundant species at the edge were Artocarpus study were of F. racemosa. heterophyllus (21.43%), Grevillia robusta (16.07%), Eucalyp- tus (14.29%), and M. eminii (10.71%; Table 3). All except

A. heterophyllus are nonnative, and are planted for timber TABLE 2.—Stand characteristics in the interior and edge of and shade in plantations. Ficus racemosa (3.57%), not native to Puduthottam, a rain-forest fragment in the Western Ghats (6 6 mean SD). the study area (although native to lowland evergreen forests), is also planted as an avenue and shade tree. Forest interior Edge When both subhabitats were pooled, M. ferrea had the Parameter (n ¼ 80) (n ¼ 56) highest basal area (6.53 m2/ha), despite the low relative density Stand density (trees/ha) 270 6 0.00 50 6 9.09 of the species (5.15%). A. heterophyllus had a basal area of Tree girth (cm) 1.12 6 0.13 1.49 6 0.14 Canopy cover (%) 94.29 6 0.67 89.74 6 2.21 4.33 m2/ha and a relative density of 10.29%. Other abundant December 2008 NANDINI AND PARTHASARATHY—DIET OF PETAURISTA IN FRAGMENTED RAIN FOREST 1553

TABLE 3.—Species composition and relative abundance (%) of plant species in the interior and edge of Puduthottam, a rain-forest fragment in the Western Ghats (n ¼ number of sampling units).

Interior Edge Species (n ¼ 20) (n ¼ 14) Maesopsis eminii Engl. (Rhamnaceae) 23.75 10.71 Dimocarpus longan Lour. (Sapindaceae) 3.75 — Diospyros (Ebenaceae) 1.25 — Casearia rubescens Dalz. (Flacourtiaceae) 3.75 — Mesua ferrea L. (Clusiaceae) 8.75 —

Macaranga () 17.5 — Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021 Phyllanthus (Euphorbiaceae) 1.25 — Holigarna nigra Bourd. (Anacardiaceae) 1.25 — Spathodea campanulata P. Beauv. (Bignoniaceae) 2.5 — Maesa indica (Roxb.) DC. (Myrsinaceae) 11.25 — hemisphericum (Wight) Alston () 1.25 — Cullenia exarillata A. Robyns () 2.50 1.79 Myristica dactyloides Gaertn. (Myristicaceae) 3.75 — Artocarpus heterophyllus Lam. () 2.5 21.43 FIG.1.—Girth-class distribution of trees in the forest interior (n ¼ Clerodendrum Vent. (Verbenaceae) 6.25 — 20 plots) and edge (n ¼ 14 plots) in Puduthottam, a rain-forest tetracoccus (Roxb.) Kurz (Euphorbiaceae) 1.25 — fragment in the Western Ghats. Dysoxylum malabaricum Bedd. ex Hiern (Meliaceae) 1.25 — Antidesma menasu (Tul.) Miq. ex Muell.-Arg. (Stilaginaceae) 1.25 — Bhesa indica (Bedd.) Ding Hou (Celastraceae) 1.25 — Leaves were consumed in 39.21% of the observations Cinnamomum sulphuratum Nees (Lauraceae) 1.25 — (mature leaves, 8.82%, and immature leaves, 7.84%). However, Cryptocarya Gamble (Lauraceae) 1.25 — the phenophase of leaves could not be ascertained in most Nothopegia racemosa (Dalz.) Ramam. (Anacardiaceae) 1.25 — observations. Leaves of F. racemosa were consumed more Toona ciliata M. Roem. (Meliaceae) — 5.36 Persea macrantha (Nees) Kosterm. (Lauraceae) — 1.79 than those of any other species (17.65%). The only flowers Alstonia scholaris (L.) R. Br. (Apocynaceae) — 3.75 observed being consumed were of Cullenia exarillata. Flying Syzygium zeylanicum P. Hermann (Myrtaceae) — 1.79 squirrels were seen eating the bark of Eucalyptus (7.84%), Grevillia robusta A. Cunn. ex R. Br. (Proteaceae) — 16.07 M. ferrea (0.98%), and Dimocarpus longan (0.98%). Ficus racemosa L. (Moraceae) — 3.57 Feeding site selection.—Comparison of the characteristics of Mangifera indica L. (Anacardiaceae) — 1.79 Eucalyptus (Myrtaceae) — 14.29 food-tree plots with random plots revealed a significant serratus L. () — 1.79 difference in total canopy contiguity (Kolmogorov–Smirnov Trichilia connaroides (Wight & Arn.) Bentvelzen 2-sample test, z ¼ 1.398, P , 0.05), with food-tree plots (Meliaceae) — 1.79 having greater canopy contiguity (6.78 trees/ha 6 0.72) than Erythrina (Fabaceae) — 7.14 random plots (2.82 trees/ha 6 0.67; Table 2). Although there Dalbergia (Fabaceae) — 3.57 Prunus ceylanica (Wight) Miq. (Rosaceae) — 1.79 was no marked preference for trees of any particular girth class (Kolmogorov–Smirnov 2-sample test, z ¼ 0.47, P . 0.05), squirrels used trees of 250–350 cm girth more than other girth classes (Fig. 3). The average height at which the squirrels were DISCUSSION observed feeding was 16.67 m and the average height of food trees was 28.75 m (n ¼ 43). During the study period (December 1999–March 2000), the Encounter rate of flying squirrels.—Flying squirrels were diet of Indian giant flying squirrels consisted chiefly of fruit sighted in coffee plantations but not in tea plantations. Coffee and leaves, although they were seen to feed on 4 different plant plantations had M. emini planted as shade cover, as well as parts. Although insectivory has been documented in flying Artocarpus, Ficus, and native rain-forest trees. Tea plantations squirrels (Muul and Lim Boo Liat 1978), we did not record had sparse tree cover, and the predominant shade tree was flying squirrels feeding on insects or animal matter during this G. robusta planted at intervals of about 20 m. Although there study. Most studies of giant flying squirrels have reported them was no significant difference in the encounter rates of flying to be largely folivorous (Kawamichi 1997; Kuo and Lee 2003; squirrels across the 3 habitats surveyed (v2 ¼ 0.581, d.f. ¼ 2, Muul and Lim Boo Liat 1978) and this is the 1st study that P . 0.05), the encounter rate seemed to be greater at the forest reports such high consumption of fruits by giant flying edge (6.1 individuals/km) and lower in the forest interior (3.92 squirrels. The fruits of F. racemosa were the most commonly individuals/km; Table 4). The encounter rate in the coffee eaten food item throughout the study. Ficus fruits are an plantations (5.62 individuals/km) may not reflect the true use of excellent source of easily assimilated energy, have low fat this habitat because most sightings in coffee plantations were content, and are a potential source of animal proteins (provided within 100 m of the forest edge. No squirrels were sighted in by the larvae of fig wasps—Vellayan 1981), calcium, and the interior regions of the plantations. minerals (O’Brien et al. 1998). 1554 JOURNAL OF MAMMALOGY Vol. 89, No. 6 Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021

FIG.3.—Girth-class distribution of food trees used by Petaurista philippensis and the available trees in Puduthottam, a rain-forest fragment, Western Ghats.

FIG.2.—Percentage observation of plant parts in the diet of Petaurista philippensis in a rain-forest fragment, Western Ghats, groups of flying squirrels were seen more often on F. racemosa during the study period (December 1999–March 2000). than on any other food source, perhaps suggesting the exploitation of figs as a major fruit source. In this study, it was ascertained that flying squirrels The encounter rates of flying squirrels in different habitats consumed more mature leaves than they did young leaves. varied and were highest at the forest edge. Increased activity of However, petioles and young leaves are known to be more species at forest edges is a well-documented phenomenon, and nutritious and less fibrous than older leaves (Coley 1983), and higher food availability was proposed as an explanation for a study of Japanese flying squirrels (Petaurista leucogenys) edge associations of Siberian flying squirrels (Pteromys used the proportion of mature leaves eaten each month as volans—Desrochers et al. 2003). At our site, the vegetation a negative index of food availability (Kawamichi 1997). The at the edge and interior differed in species composition, and the use of mature leaves during times of food scarcity also has been edge was dominated by plantation shade-cover trees. Flying observed for other folivores (Bennett and Davies 1994). In our squirrels were opportunistic in their exploitation of food study, flying squirrels ate more leaves in December and resources and more than two-thirds of their diet was composed February, when F. racemosa did not seem to fruit as of species other than native rain-forest species. The 2 most abundantly as in January. However, long-term phenological consumed species F. racemosa and A. heterophyllus, although and nutritional studies are required to supplement foraging native to the Western Ghats, occurred in unusually high studies before it can conclusively be stated that flying squirrels densities because they are planted by the cultivation estates. consumed either more figs or more mature leaves during We hypothesize that the congregation of flying squirrels at food periods of lower food availability. trees might have artificially inflated the encounter rate of flying Indian giant flying squirrels seemed to be selective in their squirrels at the edge. F. racemosa, planted along edges, was diet, like other species of Petaurista (Kawamichi 1997; Kuo clumped in distribution, and this might have contributed to its and Lee 2003), and fed only on a total of 9 plant species. exploitation by flying squirrels. Other frugivorous primates and Approximately 70% of the diet of P. philippensis during the carnivores have been observed to preferentially use large study period was composed of just 2 plant species. F. racemosa fruiting patches, because this permits individuals to feed longer was the most important food resource for flying squirrels, with in 1 patch and imposes lower travel costs (Kays 1999). Flying both fruits and leaves eaten. The role of figs as important squirrels also fed on other species that were present in good resources for frugivore communities during periods of food scarcity has been well established in rain forests (Janzen 1979; Leighton and Leighton 1983; Terborgh 1983). However, TABLE 4.—Encounter rates (number of sightings per kilometer Borges (1993) showed that figs were a lean-season resource of walk) of Petaurista philippensis in various habitat types in a only for those tree squirrels with fig sources within their fragmented landscape in the Western Ghats (December 1999–January territories, and suggested that figs can be a major fruit source 2000). only for mobile species with large home ranges. Giant flying squirrels have been observed to track resources and eat fruits Total length No. Encounter rate and flowers when available (Kuo and Lee 2003). Indian giant Habitat walked (km) sightings (no./km) flying squirrels can cover large distances by gliding (92.4 m in Forest interior 9.6 38 3.95 1 glide—R. Nandini, pers. obs.), and this would greatly Edge 40.25 246 6.11 increase their ability to track food resources. During the study, Plantations 3.375 19 5.62 December 2008 NANDINI AND PARTHASARATHY—DIET OF PETAURISTA IN FRAGMENTED RAIN FOREST 1555 numbers along the edge. Edge environments affect plant HAMILTON, A. C., AND R. BENSTEAD-SMITH (EDS.). 1990. Forest physiology, and increased light availability usually results in conservation in the East Usambara Mountains, Tanzania. Tropical greater leaf production, attracting folivores (Laurance 1991). Forestry Programme, International Union for the Conservation of An important factor affecting a species’ ability to persist at Nature and Natural Resources, Gland, Switzerland. sites after disturbance is its ability to change the relative JANZEN, D. H. 1979. How to be a fig. Annual Review of Ecology and proportion of different food types in the diet, specifically, to Systematics 10:13. JHA, C.S., C. B. S. DUTT, AND K. S. BAWA. 2000. Deforestation and exploit leaves instead of fruit (Johns 1997; Meijaard et al. land use changes in the Western Ghats, India. Current Science 2005). Barrett (1984) noted that the ability of Petaurista to 80:378–388. incorporate leaf material into its diet in times of fruit shortage JOHNS, A. D. 1985. Selective logging and wildlife conservation in greatly improved its persistence in logged forests in Malaysia. tropical rain-forest: problems and recommendations. Biological In areas disturbed by logging, squirrels have been seen to Conservation 31:355–375. Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021 respond to reduced food availability by increasing consumption JOHNS, A. G. 1997. Timber production and biodiversity conservation of bark and sap (Johns 1985). The Indian giant flying squirrel in the tropical rainforests. Cambridge University Press, Cambridge, seems to be able to exploit food resources opportunistically in United Kingdom. this disturbed site, and its long-term persistence here shows it KAWAMICHI, T. 1997. Seasonal changes in the diet of Japanese flying to be tolerant to a high amount of habitat disturbance. Further squirrels in relation to reproduction. Journal of Mammalogy studies on flying squirrel ecology and their use of resources 78:204–212. should involve determining the impact of edges and the degree KAYS, R. W. 1999. Food preferences of (Potos flavus): of isolation of the fragment on the persistence of flying squirrel a frugivorous carnivore. Journal of Mammalogy 80:589–599. populations. KREBS, C. J. 1998. Ecological methodology for Windows. Version 3.1. University of California, Berkeley. KUO, C. C., AND L. L. LEE. 2003. Food availability and food habits of ACKNOWLEDGMENTS Indian giant flying squirrels (Petaurista philippensis) in Taiwan. This study was supported by a small grant from the India Program Journal of Mammalogy 84:1330–1340. of the Wildlife Conservation Society and the Centre for Wildlife LAURANCE, W. F. 1991. Ecological correlates of extinction proneness Studies, Bangalore. The Forest Department provided in Australian tropical rain forest mammals. Conservation Biology research permits and the Manager, Puthuthottam Estate, provided 5:79–89. permission to work in this private forest. We thank D. Mudappa, LEIGHTON, M., AND D. R. LEIGHTON. 1983. Vertebrate responses to T. R. Shankar Raman, and V. V. Robin for their ideas and en- fruiting seasonality within a Bornean rain forest. Pp. 181–196 couragement through the project. We are grateful to P. Davidar in Tropical rain forest: ecology and management (S. L. Sutton, and A. Kumar for their inputs, and Silamban, Rajamani, Kumar, and T. C. Whitmore, and A. C. Chadwick, eds.). Blackwell Scientific Dinesh for assistance in the field. V. V. Robin, D. Mudappa, and Publications Ltd., London, United Kingdom. J. W. Duckworth offered useful comments on the manuscript. MEIJAARD, E., ET AL. 2005. Life after logging: reconciling wildlife conservation and production forestry in Indonesian Borneo. Center for International Forestry Research, Bogor, Indonesia. LITERATURE CITED MUUL, I., AND LIM BOO LIAT. 1978. Comparative morphology, food habits and ecology of some Malaysian arboreal rodents. Pp. 361– ASHRAF, N. V. K., A. KUMAR, AND A. J. T. JOHNSINGH. 1993. On the 368 in The ecology of arboreal folivores (G. G. Montgomery, ed.). relative abundance of two sympatric flying squirrels of Western Ghats, India. Journal of the Bombay Natural History Society Smithsonian Institution, Washington, D.C. 90:158–162. MYERS, N., R. A. MITTERMEIER,C.A.MITTERMEIER,G.A.B.DA FONSECA, AND J. KENT. 2000. Biodiversity hotspots for conservation BARRETT, E. B. M. 1984. The ecology of some nocturnal arboreal mammals in the rain forests of Peninsular Malaysia. Ph.D. dis- priorities. Nature 403:853–858. sertation, University of Cambridge, Cambridge, United Kingdom. NORUSIS, M. J. 1990. SPSS/PCþ: advanced statistics. 7.5. SPSS Inc., BENNETT, E. L., AND A. G. DAVIES. 1994. The ecology of Asian Chicago, Illinois. colobines. Pp. 129–171 in Colobine monkeys: ecology, behaviour O’BRIEN, T. G., M. F. KINNAIRD,E.S.DIERENFELD,N.L.CONKLIN- and evolution (A. G. Davies and J. F. Oates, eds.). Cambridge BRITTAIN,R.W.WRANGHAM, AND S. C. SILVER. 1998. What’s so University Press, Cambridge, United Kingdom. special about figs? Nature 392:688. BORGES, R. M. 1993. Figs, Malabar giant squirrels, and fruit shortages PASCHOAL, M., AND M. GALETTI. 1995. Seasonal food use by the within two tropical Indian forests. Biotropica 25:183–190. neotropical squirrel Sciurus ingrami in southeastern Brazil. CHAMPION, H. G., AND S. K. SETH. 1968. A revised study of the forest Biotropica 27:268–273. types of India. Government of India, New Delhi, India. PRABHAKAR, A. 1998. Impacts of habitat fragmentation on the COLEY, P. D. 1983. Herbivory and defensive characteristics of tree species terrestrial small mammal communities in the tropical rain forest in a lowland tropical forest. Ecological Monographs 53:209–233. of the Anamalai Hills in the Western Ghats, . Ph.D. CONGREVE, C. R. T. 1938. The Anamalais. Associated Printers, dissertation, Bharathidasan University, Coimbatore, India. Madras, India. RAJAMANI, N., D. MUDAPPA, AND H. VAN ROMPAEY. 2002. Distribution COTTAM, G., AND J. T. CURTIS. 1956. The use of distance methods in and status of the brown palm in the Western Ghats, South phytosociological sampling. Ecology 37:451–460. India. Small Carnivore Conservation 27:6–11. DESROCHERS, A., I. K. HANSKI, AND V. SELONEN. 2003. Siberian flying SIMBERLOFF, D., AND T. DAYAN. 1991. Guilds and the structure of squirrel responses to high- and low-contrast forest edges. Landscape ecological communities. Annual Review of Ecology and System- Ecology 18:543–552. atics 22:115–143. 1556 JOURNAL OF MAMMALOGY Vol. 89, No. 6

TERBORGH, J. 1983. Five new world primates. Princeton University WEIGL, P. D., T. W. KNOWLES, AND A. C. BOUYTON. 1999. The Press, Princeton, New Jersey. distribution and ecology of the northern flying squirrel, Glaucomys THIOLLAY, J. M. 1992. Influence of selective logging on bird species sabrinus coloratus, in the southern Appalachians. North Carolina diversity in a Guianan rain forest. Conservation Biology 6: Wildlife Resources Commission, Non-game and Endangered 47–63. Wildlife Program, Division of Wildlife Management, Raleigh, UMAPATHY, G. U., AND A. KUMAR. 2000. The occurrence of arboreal North Carolina. mammals in the rain forest fragments in the Anamalai Hills, South India. Biological Conservation 92:311–319. Submitted 15 February 2008. Accepted 21 May 2008. VELLAYAN, S. 1981. The nutritive value of Ficus in the diet of lar gibbon (Hylobates lar). Malaysian Applied Biology 10:177–181. Associate Editor was Rodrigo A. Medellı´n. Downloaded from https://academic.oup.com/jmammal/article/89/6/1550/911817 by guest on 28 September 2021